Comparison of precipitation measurements by OTT Parsivel2 and Thies LPM optical disdrometers
Analysis of minute and event disdrometer data, corresponding to the article of the same title in Hydrology and Earth System Sciences
Santiago Beguería
28 febrero, 2018
1 Introduction and dataset
We analyze one-minute disdrometric data recorded by two Thies LPM and two Parsivel2 disdrometers at the Aula Dei Experimental Station (41º43’30”N, 0º48’39”W, 230 m.a.s.l.), between 2013-06-17 and 2015-07-21. The disdrometers where installed on two masts with a separation of 1.5 m between them (Figure 1).
A total of 510 events were recorded during that period, spanning between 5 and 1454 minutes. This sums up to ~75k to ~99k one-minute records, depending on the device (see main article, Table 3).
The dataset analysed here corresponds to the common minutes, defined as those having high quality data and detection of rainfall particles in each of the four disdrometers (see main article, secion 2.3). This led to a total of 46,636 records, corresponding to 11,659 minutes belonging to 157 rainfall episodes.
We shall start by reading and formatting the dataset. There are two alternative datasets: * file ./data/data.csv.gz, with variables as measured by devices and calculated from the particle size and velocity (PSVD) data, after a filtering and correction procedure was applied to the raw PSVD data to remove ulikely particle size and velocity bins (see main article, section 2.2); * file ./data/data_unfiltered.csv.gz, as the previous one but using the raw PSVD data (i.e., with no filtering and correction).
The first dataset will be stored in object ‘dat’, while the second dataset will be stored in object ‘datunf’.
# load data (filtered dataset)
dat <- read.table('./data/data.csv.gz', sep=',', head=TRUE)
# format factors
dat$Event <- as.factor(dat$Event)
dat$ID <- factor(dat$ID, levels=c('T1','T2','P1','P2'))
dat$Type <- factor(dat$Type)
# format time
dat$Time <- strptime(dat$Time, format='%Y-%m-%d %H:%M:%S')
head(dat)## Time Event ID Serial Type Mast NP_meas R_meas Z_meas
## 1 2013-06-17 11:42:00 2 T1 436 Thi 1 121 1.400 30.8
## 2 2013-06-17 11:43:00 2 T1 436 Thi 1 96 0.921 27.1
## 3 2013-06-17 11:56:00 2 T1 436 Thi 1 26 0.275 20.7
## 4 2013-06-17 23:31:00 3 T1 436 Thi 1 27 0.330 20.8
## 5 2013-06-17 23:32:00 3 T1 436 Thi 1 29 0.355 21.3
## 6 2013-06-17 23:33:00 3 T1 436 Thi 1 47 0.506 22.4
## E_meas Pcum_meas Ecum_meas NP ND R P Z M E
## 1 NA 0.02333333 NA 71 7265.362 1.3860 0.023110 31.49 0 22.07
## 2 NA 0.03868333 NA 69 6871.656 0.9138 0.015230 27.06 0 16.59
## 3 NA 0.04326667 NA 22 6480.489 0.3137 0.005228 22.16 0 14.96
## 4 NA 0.04876667 NA 21 3602.272 0.3440 0.005733 21.17 0 13.43
## 5 NA 0.05468333 NA 26 4801.632 0.3595 0.005992 21.63 0 14.20
## 6 NA 0.06311667 NA 38 3996.339 0.5613 0.009355 23.93 0 13.54
## Pcum Ecum D10 D25 D50 D75 D90 Dm V10 V25
## 1 0.02310000 0.3678333 0.262 0.296 0.351 0.657 1.496 0.858 0.871 0.996
## 2 0.03833000 0.6443333 0.321 0.366 0.886 1.239 1.408 0.990 1.035 1.303
## 3 0.04355833 0.8936667 0.318 0.334 0.684 1.133 1.343 0.986 1.140 1.316
## 4 0.04929167 1.1175000 0.370 0.692 1.026 1.369 1.492 1.195 1.386 4.146
## 5 0.05528333 1.3541667 0.294 0.701 1.089 1.173 1.260 1.113 1.329 3.390
## 6 0.06463833 1.5798333 0.341 0.793 1.038 1.143 1.280 1.162 1.555 3.320
## V50 V75 V90 Vm
## 1 1.284 4.153 4.983 2.795
## 2 3.937 4.581 4.960 3.602
## 3 3.448 4.252 5.007 3.663
## 4 4.384 4.720 4.990 4.516
## 5 3.816 4.200 4.600 4.075
## 6 3.828 4.549 4.873 4.233# load data (unfiltered dataset)
datunf <- read.table('./data/data_unfiltered.csv.gz', sep=',', head=TRUE)
# format factors
datunf$Event <- as.factor(datunf$Event)
datunf$ID <- factor(datunf$ID, levels=c('T1','T2','P1','P2'))
datunf$Type <- factor(datunf$Type)
# format time
datunf$Time <- strptime(datunf$Time, format='%Y-%m-%d %H:%M:%S')
head(datunf)## Time Event ID Serial Type Mast NP_meas R_meas Z_meas
## 1 2013-06-17 11:42:00 2 T1 436 Thi 1 121 1.400 30.8
## 2 2013-06-17 11:43:00 2 T1 436 Thi 1 96 0.921 27.1
## 3 2013-06-17 11:56:00 2 T1 436 Thi 1 26 0.275 20.7
## 4 2013-06-17 23:31:00 3 T1 436 Thi 1 27 0.330 20.8
## 5 2013-06-17 23:32:00 3 T1 436 Thi 1 29 0.355 21.3
## 6 2013-06-17 23:33:00 3 T1 436 Thi 1 47 0.506 22.4
## E_meas Pcum_meas Ecum_meas NP ND R P Z M E
## 1 NA 0.02333333 NA 121 17439.908 1.3130 0.021880 31.22 0 21.94
## 2 NA 0.03868333 NA 96 16015.665 0.8768 0.014610 26.82 0 16.45
## 3 NA 0.04326667 NA 26 9505.014 0.3006 0.005011 21.95 0 14.92
## 4 NA 0.04876667 NA 27 5673.157 0.3338 0.005563 21.00 0 13.41
## 5 NA 0.05468333 NA 29 6203.553 0.3461 0.005769 21.43 0 14.12
## 6 NA 0.06311667 NA 47 7900.304 0.5408 0.009013 23.73 0 13.48
## Pcum Ecum D10 D25 D50 D75 D90 Dm V10 V25
## 1 0.02188333 0.3656667 0.167 0.212 0.298 0.373 1.846 0.574 0.671 0.881
## 2 0.03649667 0.6398333 0.181 0.249 0.452 1.231 1.515 0.756 0.596 1.046
## 3 0.04150667 0.8885000 0.210 0.296 0.587 1.378 1.632 0.856 1.016 1.110
## 4 0.04707000 1.1120000 0.307 0.455 1.085 1.525 1.637 1.007 1.159 3.807
## 5 0.05283833 1.3473333 0.296 0.622 1.142 1.342 1.525 1.017 1.117 3.489
## 6 0.06185167 1.5720000 0.166 0.362 1.063 1.354 1.587 0.961 0.959 3.279
## V50 V75 V90 Vm
## 1 1.152 2.008 5.235 2.061
## 2 3.356 4.500 5.648 2.878
## 3 3.856 5.056 5.684 3.246
## 4 4.424 5.082 5.795 4.322
## 5 4.037 4.582 5.546 3.893
## 6 4.100 4.698 5.492 3.777These are the variables included in the datasets:
| var | full name | units |
|---|---|---|
| Time | time of the record | Y-m-d hh:mm:ss |
| Event | event ID | (factor) |
| ID | disdromter ID | (factor: T1, T2, P1, P2) |
| Serial | disdrometer serial number | (factor) |
| Type | disdrometer type | (factor: Thi, Par) |
| Mast | mast ID | (factor: 1, 2) |
| NP_meas | number of particles detected | (-) |
| R_meas | rainfall intensity, as outputted by the device | \(mm\ h^{-1}\) |
| Z_meas | radar reflectivity, as outputted by the device | \(dB\ mm^6\ m^{-3}\) |
| E_meas | erosivity, as outputted by the device | \(J\ m^{-2}\ mm^{-1}\) |
| Pcum_meas | cumulative rainfall amount | \(mm\) |
| Ecum_meas | cumulative kinetic energy | \(J\ m^{-2}\ mm^{-1}\) |
| NP | number of particles detected | (-) |
| ND | particle density | \(m^{-3}\ mm^{-1}\) |
| R | rainfall intensity | \(mm\ h^{-1}\) |
| P | rainfall amount | \(mm\) |
| Z | radar reflectivity | \(dB\ mm^6\ m^{-3}\) |
| M | water content | \(g m^{-3}\) |
| E | kinetic energy | \(J\ m^{-2}\ mm^{-1}\) |
| Pcum | cumulative rainfall amount | \(mm\) |
| Ecum | cumulative kinetic energy | \(J\ m^{-2}\ mm^{-1}\) |
| D10 | drop diameter’s 10th percentile | \(mm\) |
| D25 | drop diameter’s 25th percentile | \(mm\) |
| D50 | drop diameter’s 50th percentile | \(mm\) |
| D75 | drop diameter’s 75th percentile | \(mm\) |
| D90 | drop diameter’s 90th percentile | \(mm\) |
| Dm | mean drop diameter | \(mm\) |
| V10 | drop velocity’s 10th percentile | \(m\ s^{-1}\) |
| V25 | drop velocity’s 25th percentile | \(m\ s^{-1}\) |
| V50 | drop velocity’s 50th percentile | \(m\ s^{-1}\) |
| V75 | drop velocity’s 75th percentile | \(m\ s^{-1}\) |
| V90 | drop velocity’s 90th percentile | \(m\ s^{-1}\) |
| Vm | mean drop velocity | \(m\ s^{-1}\) |
2 Exploratory analysis
With the objective of comparing the behaviour of the two disdrometer types for several variables, we use different plotting tools for a first, exploratory, analysis. Different colors and linetypes will be used to discriminate between measuring devices.
2.1 Cumulative variables
Time series of cumulative precipitation and kinetic energy over the whole experiment. The internally computed values (measured) are compared to those calculated from the PSVD (calculated). We will produce plots for both filtered and unfiltered data.
Filtered data
Unfiltered data
2.2 Density plots, all minutes
We will now have a look at the kernel densities of the integrated variables. All the minute records will be included in the plot.
Filtered data
Unfiltered data
2.3 Violin plots
The same, but with violin plots (might be easier to read to some),
Filtered data
Unfiltered data
2.4 Density plots, by intensity ranges
We will repeat now the kernel density plots, but discriminating between low, medium and high precipitation intensity.
0.1 mm/h < R < 2 mm/h; N = 52306
2 mm/h < R < 10 mm/h; N = 6954
R > 10 mm/h; N = 1377
2.5 Density plots, events
We will now make density plots as the previous ones, but for events (and not minutes).
3 Analysis
We will use a Gamma Generalized Linear Mixed-Effects Model to compare between the two disdrometer types, with the factor Type as fixed effect and the factor Mast as random variable. For several dependet variables (ND, R, E) we use a log link, while we use an identity link for the rest (Z, D10, D50, etc).
3.1 Minute data, all intensities
# take a random sample of 250 complete minutes x 4 devices = 1000
set.seed(12345)
smp <- names(which(table(as.character(dat$Time))==4))
smp <- sample(smp, 250) # times 4 = 1000 records
smp <- which(dat$Time %in% smp)
m_np <- glmer(NP ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_np)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: NP ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 12597.1 12616.7 -6294.6 12589.1 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0821 -0.6786 -0.2859 0.3537 7.1889
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0000 0.0000
## Residual 0.7602 0.8719
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 5.26687 0.03494 150.7 <2e-16 ***
## TypeThi 5.43832 0.03494 155.6 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_np)$coeff[,1]),4)## TypePar TypeThi
## 193.8 230.1m_d10 <- glmer(D10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -2682.4 -2662.8 1345.2 -2690.4 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2831 -0.6322 -0.1682 0.3860 7.6834
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 1.306e-05 0.003614
## Residual 2.993e-02 0.173013
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.477206 0.005418 88.08 <2e-16 ***
## TypeThi 0.337419 0.004865 69.36 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.682signif(summary(m_d10)$coeff[,1],4)## TypePar TypeThi
## 0.4772 0.3374m_d50 <- glmer(D50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -1424.3 -1404.7 716.2 -1432.3 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.6106 -0.6264 -0.0944 0.4341 6.9348
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 2.215e-06 0.001488
## Residual 3.607e-02 0.189923
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.742047 0.006127 121.1 <2e-16 ***
## TypeThi 0.595601 0.004990 119.4 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.066signif(summary(m_d50)$coeff[,1],4)## TypePar TypeThi
## 0.7420 0.5956m_d90 <- glmer(D90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -374.0 -354.4 191.0 -382.0 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4395 -0.7197 -0.1848 0.5361 5.1552
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.04363 0.2089
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.025600 0.009111 112.6 <2e-16 ***
## TypeThi 1.011874 0.008989 112.6 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_d90)$coeff[,1],4)## TypePar TypeThi
## 1.026 1.012m_v10 <- glmer(V10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 464.6 484.2 -228.3 456.6 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.5246 -0.6805 -0.1931 0.5112 6.2395
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0002945 0.01716
## Residual 0.0439691 0.20969
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.79271 0.02637 67.99 <2e-16 ***
## TypeThi 1.31562 0.02401 54.80 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.691signif(summary(m_v10)$coeff[,1],4)## TypePar TypeThi
## 1.793 1.316m_v50 <- glmer(V50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 1127.6 1147.2 -559.8 1119.6 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.9613 -0.6035 -0.0436 0.4903 4.7745
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0006002 0.0245
## Residual 0.0270815 0.1646
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.87503 0.03803 75.59 <2e-16 ***
## TypeThi 2.39933 0.03625 66.18 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.733signif(summary(m_v50)$coeff[,1],4)## TypePar TypeThi
## 2.875 2.399m_v90 <- glmer(V90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 1520.7 1540.3 -756.4 1512.7 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.8928 -0.6610 -0.1318 0.4945 5.4572
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.000144 0.0120
## Residual 0.020889 0.1445
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.60780 0.02563 140.8 <2e-16 ***
## TypeThi 3.81758 0.02679 142.5 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.217signif(summary(m_v90)$coeff[,1],4)## TypePar TypeThi
## 3.608 3.818m_nd <- glmer(ND ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_nd)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: ND ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 21215.0 21234.6 -10603.5 21207.0 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6431 -0.7528 -0.2152 0.5783 3.4818
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.000 0.000
## Residual 0.334 0.578
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 9.67556 0.02666 362.9 <2e-16 ***
## TypeThi 9.98055 0.02666 374.3 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_nd)$coeff[,1]),4)## TypePar TypeThi
## 15920 21600m_r <- glmer(R ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_r)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: R ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 2611.0 2630.7 -1301.5 2603.0 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.6338 -0.4918 -0.3096 0.0853 11.2450
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 5.252e-16 2.292e-08
## Residual 2.153e+00 1.467e+00
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.22639 0.04325 5.234 1.66e-07 ***
## TypeThi 0.36455 0.04325 8.428 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_r)$coeff[,1]),4)## TypePar TypeThi
## 1.254 1.440m_z <- glmer(Z ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_z)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: Z ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 6629.8 6649.5 -3310.9 6621.8 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.1169 -0.7365 -0.0692 0.6558 4.6984
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.07999 0.2828
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.14564 0.01275 246.7 <2e-16 ***
## TypeThi 3.20075 0.01275 251.0 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_z)$coeff[,1]),4)## TypePar TypeThi
## 23.23 24.55m_e <- glmer(E ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_e)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: E ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 5718.0 5737.7 -2855.0 5710.0 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.3507 -0.6936 -0.2482 0.3910 6.4739
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 4.406e-16 2.099e-08
## Residual 2.413e-01 4.912e-01
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.26802 0.01936 117.1 <2e-16 ***
## TypeThi 2.40603 0.01936 124.3 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_e)$coeff[,1]),4)## TypePar TypeThi
## 9.66 11.09# pars <- rbind(c(summary(m_r)$co[c(1,2),1],summary(m_r)$co[1,2]),
# c(summary(m_e)$co[c(1,2),1],summary(m_e)$co[1,2]))
# rownames(pars) <- c('r','e')
# kable(pars, digits=4,
# col.names=c('Parsivel','Thies','st. error'))3.2 Minute data, low intensities
set.seed(12345)
w <- dat$R>0.1 & dat$R<2
smp <- names(which(table(as.character(dat$Time[w]))==4))
smp <- sample(smp, 250)
smp <- which(dat$Time %in% smp)
m_np <- glmer(NP ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_np)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: NP ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 11611.7 11631.4 -5801.9 11603.7 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.2872 -0.6908 -0.2998 0.4176 5.5360
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 1.282e-14 1.132e-07
## Residual 5.082e-01 7.129e-01
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 4.91330 0.02991 164.3 <2e-16 ***
## TypeThi 4.98192 0.02991 166.6 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_np)$coeff[,1]),4)## TypePar TypeThi
## 136.1 145.8m_d10 <- glmer(D10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -2631.5 -2611.9 1319.8 -2639.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3397 -0.6646 -0.1859 0.3413 5.9460
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 1.594e-05 0.003992
## Residual 2.939e-02 0.171445
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.474142 0.006038 78.52 <2e-16 ***
## TypeThi 0.346188 0.005572 62.13 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.739signif(summary(m_d10)$coeff[,1],4)## TypePar TypeThi
## 0.4741 0.3462m_d50 <- glmer(D50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -1758.7 -1739.0 883.3 -1766.7 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.6658 -0.5949 -0.1181 0.4507 7.3972
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 1.89e-05 0.004348
## Residual 2.65e-02 0.162801
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.711337 0.006915 102.88 <2e-16 ***
## TypeThi 0.596354 0.006374 93.56 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.536signif(summary(m_d50)$coeff[,1],4)## TypePar TypeThi
## 0.7113 0.5964m_d90 <- glmer(D90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -1048.6 -1029.0 528.3 -1056.6 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.1166 -0.7311 -0.1326 0.5222 6.3904
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.02441 0.1562
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.956236 0.006453 148.2 <2e-16 ***
## TypeThi 0.947202 0.006392 148.2 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_d90)$coeff[,1],4)## TypePar TypeThi
## 0.9562 0.9472m_v10 <- glmer(V10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 566.7 586.3 -279.3 558.7 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.1238 -0.6849 -0.1753 0.4909 5.3177
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0004318 0.02078
## Residual 0.0473658 0.21764
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.79182 0.03143 57.02 <2e-16 ***
## TypeThi 1.35611 0.02948 46.00 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.767signif(summary(m_v10)$coeff[,1],4)## TypePar TypeThi
## 1.792 1.356m_v50 <- glmer(V50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 902.5 922.1 -447.3 894.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.2335 -0.6208 -0.0640 0.4951 4.8059
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0006181 0.02486
## Residual 0.0222395 0.14913
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.79649 0.03875 72.17 <2e-16 ***
## TypeThi 2.40451 0.03760 63.95 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.800signif(summary(m_v50)$coeff[,1],4)## TypePar TypeThi
## 2.796 2.405m_v90 <- glmer(V90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 912.3 932.0 -452.2 904.3 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.8084 -0.6630 -0.0925 0.5484 6.3241
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0002455 0.01567
## Residual 0.0118714 0.10896
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.45799 0.02462 140.5 <2e-16 ***
## TypeThi 3.65312 0.02526 144.6 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.532signif(summary(m_v90)$coeff[,1],4)## TypePar TypeThi
## 3.458 3.653m_r <- glmer(R ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_r)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: R ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 832.1 851.8 -412.1 824.1 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.2226 -0.8130 -0.3133 0.5859 3.2387
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0000 0.0000
## Residual 0.4785 0.6917
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar -0.49609 0.03022 -16.42 <2e-16 ***
## TypeThi -0.40135 0.03022 -13.28 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_r)$coeff[,1]),4)## TypePar TypeThi
## 0.6089 0.6694m_nd <- glmer(ND ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_nd)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: ND ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 21246.3 21265.9 -10619.1 21238.3 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8728 -0.7468 -0.1666 0.6511 4.1239
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 5.026e-15 7.090e-08
## Residual 2.564e-01 5.064e-01
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 9.79288 0.02398 408.5 <2e-16 ***
## TypeThi 10.05533 0.02398 419.4 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_nd)$coeff[,1]),4)## TypePar TypeThi
## 17910 23280m_z <- glmer(Z ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_z)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: Z ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 6090.3 6110.0 -3041.2 6082.3 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.41988 -0.78343 -0.04816 0.75447 2.65707
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.05775 0.2403
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.01534 0.01098 274.7 <2e-16 ***
## TypeThi 3.07694 0.01098 280.3 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_z)$coeff[,1]),4)## TypePar TypeThi
## 20.40 21.69m_e <- glmer(E ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_e)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: E ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 5251.7 5271.4 -2621.9 5243.7 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.3577 -0.7019 -0.2914 0.4453 7.9470
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 7.157e-05 0.00846
## Residual 1.995e-01 0.44671
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.10533 0.01891 111.3 <2e-16 ***
## TypeThi 2.28675 0.01891 120.9 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.150signif(exp(summary(m_e)$coeff[,1]),4)## TypePar TypeThi
## 8.210 9.8433.3 Minute data, medium intensities
set.seed(12345)
w <- dat$R>2 & dat$R<10
smp <- names(which(table(as.character(dat$Time[w]))==4))
smp <- sample(smp, 250)
smp <- which(dat$Time %in% smp)
m_np <- glmer(NP ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_np)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: NP ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 13197.5 13217.1 -6594.7 13189.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.1618 -0.6774 -0.1187 0.5470 4.4012
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 9.887e-18 3.144e-09
## Residual 1.611e-01 4.014e-01
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 6.01151 0.01818 330.7 <2e-16 ***
## TypeThi 6.25221 0.01818 343.9 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_np)$coeff[,1]),4)## TypePar TypeThi
## 408.1 519.2m_d10 <- glmer(D10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -3048.5 -3028.9 1528.3 -3056.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.6541 -0.5325 -0.2016 0.3660 8.7598
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 9.193e-07 0.0009588
## Residual 1.988e-02 0.1410025
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.499927 0.003135 159.4 <2e-16 ***
## TypeThi 0.311117 0.002086 149.2 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.135signif(summary(m_d10)$coeff[,1],4)## TypePar TypeThi
## 0.4999 0.3111m_d50 <- glmer(D50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -1410.3 -1390.6 709.1 -1418.3 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.5283 -0.5522 -0.0745 0.5393 6.0584
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 2.666e-19 5.163e-10
## Residual 3.057e-02 1.748e-01
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.835236 0.006410 130.3 <2e-16 ***
## TypeThi 0.586644 0.004502 130.3 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_d50)$coeff[,1],4)## TypePar TypeThi
## 0.8352 0.5866m_d90 <- glmer(D90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -764.4 -744.8 386.2 -772.4 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.3493 -0.6430 -0.1135 0.5268 5.4199
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 3.219e-05 0.005673
## Residual 1.790e-02 0.133799
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.289742 0.009674 133.3 <2e-16 ***
## TypeThi 1.262245 0.009551 132.2 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.409signif(summary(m_d90)$coeff[,1],4)## TypePar TypeThi
## 1.290 1.262m_v10 <- glmer(V10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 126.1 145.7 -59.0 118.1 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.7366 -0.5881 -0.1272 0.4745 5.0423
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 8.311e-05 0.009116
## Residual 3.081e-02 0.175517
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.91711 0.01802 106.38 <2e-16 ***
## TypeThi 1.16826 0.01366 85.55 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.425signif(summary(m_v10)$coeff[,1],4)## TypePar TypeThi
## 1.917 1.168m_v50 <- glmer(V50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 1302.1 1321.8 -647.1 1294.1 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.0899 -0.4481 0.0094 0.4225 3.8752
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0005477 0.0234
## Residual 0.0276997 0.1664
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.16341 0.03854 82.08 <2e-16 ***
## TypeThi 2.38821 0.03519 67.87 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.672signif(summary(m_v50)$coeff[,1],4)## TypePar TypeThi
## 3.163 2.388m_v90 <- glmer(V90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 1137.4 1157.0 -564.7 1129.4 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.7301 -0.6187 -0.1094 0.4792 6.0437
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0008789 0.02965
## Residual 0.0105450 0.10269
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 4.16510 0.04485 92.87 <2e-16 ***
## TypeThi 4.48776 0.04545 98.74 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.822signif(summary(m_v90)$coeff[,1],4)## TypePar TypeThi
## 4.165 4.488m_r <- glmer(R ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_r)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: R ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 3350.0 3369.6 -1671.0 3342.0 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.2973 -0.7496 -0.3104 0.4944 3.6479
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0004825 0.02197
## Residual 0.1498136 0.38706
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.24730 0.02908 42.90 <2e-16 ***
## TypeThi 1.40275 0.02908 48.23 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.700signif(exp(summary(m_r)$coeff[,1]),4)## TypePar TypeThi
## 3.481 4.066m_nd <- glmer(ND ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_nd)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: ND ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 19557.2 19576.9 -9774.6 19549.2 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2685 -0.6657 -0.0538 0.5041 4.7304
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0003262 0.01806
## Residual 0.1686467 0.41067
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 8.99852 0.02821 319.0 <2e-16 ***
## TypeThi 9.51987 0.02822 337.3 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.532signif(exp(summary(m_nd)$coeff[,1]),4)## TypePar TypeThi
## 8091 13630m_z <- glmer(Z ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_z)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: Z ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 5300.8 5320.5 -2646.4 5292.8 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3300 -0.6868 -0.0350 0.6187 6.2366
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 6.607e-05 0.008128
## Residual 1.070e-02 0.103450
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.47242 0.01343 258.5 <2e-16 ***
## TypeThi 3.54035 0.01343 263.5 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.884signif(exp(summary(m_z)$coeff[,1]),4)## TypePar TypeThi
## 32.21 34.48m_e <- glmer(E ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_e)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: E ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 5989.7 6009.3 -2990.8 5981.7 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.7965 -0.6786 -0.1749 0.4920 7.3097
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0002245 0.01498
## Residual 0.1337879 0.36577
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.65701 0.02166 122.7 <2e-16 ***
## TypeThi 2.72613 0.02165 125.9 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.510signif(exp(summary(m_e)$coeff[,1]),4)## TypePar TypeThi
## 14.25 15.273.4 Minute data, high intensities
set.seed(12345)
w <- dat$R>10
smp <- names(which(table(as.character(dat$Time[w]))==4))
#smp <- sample(smp, 250)
smp <- which(dat$Time %in% smp)
m_np <- glmer(NP ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_np)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: NP ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 2062.0 2073.8 -1027.0 2054.0 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.79405 -0.70194 -0.07279 0.70325 2.79076
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 8.581e-17 9.263e-09
## Residual 1.247e-01 3.532e-01
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 6.72105 0.04327 155.3 <2e-16 ***
## TypeThi 7.22026 0.04327 166.9 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_np)$coeff[,1]),4)## TypePar TypeThi
## 829.7 1367.0m_d10 <- glmer(D10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -322.4 -310.7 165.2 -330.4 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2763 -0.3532 -0.0961 0.2916 4.3168
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.03624 0.1904
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.541771 0.012262 44.18 <2e-16 ***
## TypeThi 0.287257 0.006502 44.18 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_d10)$coeff[,1],4)## TypePar TypeThi
## 0.5418 0.2873m_d50 <- glmer(D50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## -53.8 -42.0 30.9 -61.8 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8435 -0.6324 -0.1969 0.5777 4.3107
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.08057 0.2839
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.02159 0.03329 30.68 <2e-16 ***
## TypeThi 0.51403 0.01675 30.68 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_d50)$coeff[,1],4)## TypePar TypeThi
## 1.022 0.514m_d90 <- glmer(D90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_d90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 4.7 16.4 1.7 -3.3 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.8891 -0.5761 0.0475 0.6547 2.0580
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0003012 0.01735
## Residual 0.0192539 0.13876
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.75368 0.03625 48.38 <2e-16 ***
## TypeThi 1.54184 0.03323 46.40 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.323signif(summary(m_d90)$coeff[,1],4)## TypePar TypeThi
## 1.754 1.542m_v10 <- glmer(V10 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V10 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 93.9 105.7 -43.0 85.9 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.3019 -0.3373 -0.0704 0.3722 3.2634
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.000
## Residual 0.05018 0.224
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.05021 0.05654 36.26 <2e-16 ***
## TypeThi 1.01251 0.02792 36.26 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_v10)$coeff[,1],4)## TypePar TypeThi
## 2.050 1.013m_v50 <- glmer(V50 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V50 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 303.6 315.4 -147.8 295.6 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8863 -0.6943 -0.1595 0.5437 3.3335
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 3.003e-15 5.48e-08
## Residual 7.509e-02 2.74e-01
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.51254 0.11130 31.56 <2e-16 ***
## TypeThi 2.02566 0.06419 31.56 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_v50)$coeff[,1],4)## TypePar TypeThi
## 3.513 2.026m_v90 <- glmer(V90 ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='identity'))
summary(m_v90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V90 ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 251.0 262.8 -121.5 243.0 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.7136 -0.5556 0.1160 0.7763 2.3175
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.01136 0.1066
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 5.42417 0.07127 76.1 <2e-16 ***
## TypeThi 5.03647 0.06618 76.1 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_v90)$coeff[,1],4)## TypePar TypeThi
## 5.424 5.036m_r <- glmer(R ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_r)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: R ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 794.0 805.8 -393.0 786.0 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.1622 -0.7134 -0.2855 0.3048 3.7446
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0006211 0.02492
## Residual 0.0824078 0.28707
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.67881 0.04016 66.70 <2e-16 ***
## TypeThi 2.79071 0.04018 69.45 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.385signif(exp(summary(m_r)$coeff[,1]),4)## TypePar TypeThi
## 14.57 16.29m_nd <- glmer(ND ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_nd)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: ND ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 2593.9 2605.7 -1293.0 2585.9 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8497 -0.7144 -0.1017 0.6164 2.9441
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0000 0.000
## Residual 0.1772 0.421
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 8.19551 0.05125 159.9 <2e-16 ***
## TypeThi 9.24373 0.05125 180.4 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_nd)$coeff[,1]),4)## TypePar TypeThi
## 3625 10340m_z <- glmer(Z ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_z)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: Z ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 780.4 792.2 -386.2 772.4 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.6260 -0.7544 -0.1384 0.6452 3.4314
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0001039 0.01019
## Residual 0.0083872 0.09158
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.71113 0.01608 230.8 <2e-16 ***
## TypeThi 3.76472 0.01608 234.1 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.552signif(exp(summary(m_z)$coeff[,1]),4)## TypePar TypeThi
## 40.90 43.15m_e <- glmer(E ~ 0 + Type + (1|Mast), data=dat[smp,], Gamma(link='log'))
summary(m_e)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: E ~ 0 + Type + (1 | Mast)
## Data: dat[smp, ]
##
## AIC BIC logLik deviance df.resid
## 921.7 933.5 -456.9 913.7 136
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.67339 -0.78135 -0.02104 0.72178 2.41577
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0000 0.0000
## Residual 0.1005 0.3169
## Number of obs: 140, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.02246 0.03853 78.44 <2e-16 ***
## TypeThi 2.98655 0.03853 77.51 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_e)$coeff[,1]),4)## TypePar TypeThi
## 20.54 19.823.5 Event data
set.seed(12345)
smp <- names(which(table(as.character(events$ev))==4))
smp <- which(events$ev %in% smp)
events$Type <- as.factor(substr(events$ID, 1, 1))
events$Mast <- as.factor(substr(events$ID, 2, 2))
m_np <- glmer(NPm ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='log'))
summary(m_np)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: NPm ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 7445.9 7463.7 -3719.0 7437.9 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.0900 -0.6855 -0.2911 0.4095 6.4404
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0000 0.0000
## Residual 0.7162 0.8463
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 4.98561 0.04127 120.8 <2e-16 ***
## TypeT 5.12095 0.04127 124.1 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(exp(summary(m_np)$coeff[,1]),4)## TypeP TypeT
## 146.3 167.5m_d10m <- glmer(D10m ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_d10m)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D10m ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## -1703.6 -1685.9 855.8 -1711.6 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2966 -0.5831 -0.1623 0.2727 7.1333
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 9.440e-06 0.003073
## Residual 2.653e-02 0.162887
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 0.490417 0.005284 92.82 <2e-16 ***
## TypeT 0.344783 0.004375 78.81 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.457signif(summary(m_d10m)$coeff[,1],4)## TypeP TypeT
## 0.4904 0.3448m_d50m <- glmer(D50m ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_d50m)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D50m ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## -1094.1 -1076.3 551.0 -1102.1 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2411 -0.5889 -0.1982 0.4353 5.1513
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.02446 0.1564
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 0.755960 0.006363 118.8 <2e-16 ***
## TypeT 0.606134 0.005102 118.8 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(summary(m_d50m)$coeff[,1],4)## TypeP TypeT
## 0.7560 0.6061m_d90m <- glmer(D90m ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_d90m)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D90m ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## -573.2 -555.4 290.6 -581.2 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.1091 -0.6793 -0.1957 0.5421 4.3635
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.02453 0.1566
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 1.027353 0.008781 117 <2e-16 ***
## TypeT 0.997098 0.008522 117 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(summary(m_d90m)$coeff[,1],4)## TypeP TypeT
## 1.0270 0.9971m_v10 <- glmer(V10m ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_v10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V10m ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 246.6 264.3 -119.3 238.6 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.9823 -0.5803 -0.1328 0.3729 7.5876
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0004109 0.02027
## Residual 0.0414589 0.20361
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 1.82617 0.03047 59.94 <2e-16 ***
## TypeT 1.35088 0.02751 49.10 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.656signif(summary(m_v10)$coeff[,1],4)## TypeP TypeT
## 1.826 1.351m_v50m <- glmer(V50m ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_v50m)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V50m ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 490.1 507.8 -241.0 482.1 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.8147 -0.6161 -0.1298 0.4716 4.0429
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0006798 0.02607
## Residual 0.0189075 0.13750
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 2.87634 0.04004 71.83 <2e-16 ***
## TypeT 2.46546 0.03843 64.16 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.732signif(summary(m_v50m)$coeff[,1],4)## TypeP TypeT
## 2.876 2.465m_v90m <- glmer(V90m ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_v90m)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V90m ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 641.5 659.2 -316.7 633.5 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4097 -0.6501 -0.1493 0.5465 4.4468
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0003638 0.01907
## Residual 0.0124082 0.11139
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 3.59669 0.03087 116.5 <2e-16 ***
## TypeT 3.79053 0.03174 119.4 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.469signif(summary(m_v90m)$coeff[,1],4)## TypeP TypeT
## 3.597 3.791m_rm <- glmer(Rm ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_rm)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: Rm ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 1213.5 1231.2 -602.8 1205.5 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.8291 -0.6023 -0.3136 0.1326 6.8324
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.000 0.000
## Residual 1.152 1.073
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 0.96152 0.04431 21.7 <2e-16 ***
## TypeT 1.05067 0.04842 21.7 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(summary(m_rm)$coeff[,1],4)## TypeP TypeT
## 0.9615 1.0510m_rmx <- glmer(RM ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_rmx)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: RM ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 2769.6 2787.4 -1380.8 2761.6 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.6106 -0.5117 -0.3353 0.0196 7.6037
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 4.607e-15 6.788e-08
## Residual 2.508e+00 1.584e+00
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 3.4303 0.2155 15.92 <2e-16 ***
## TypeT 3.3511 0.2105 15.92 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(summary(m_rmx)$coeff[,1],4)## TypeP TypeT
## 3.430 3.351m_nd <- glmer(NDm ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_nd)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: NDm ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 13049.2 13067.0 -6520.6 13041.2 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.9160 -0.7685 -0.0458 0.6451 6.7750
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 8.618e-09 9.283e-05
## Residual 2.223e-01 4.714e-01
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 15930.75 24.19 658.6 <2e-16 ***
## TypeT 20775.60 94.53 219.8 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(summary(m_nd)$coeff[,1],4)## TypeP TypeT
## 15930 20780m_em <- glmer(Em ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_em)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: Em ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 3255.4 3273.1 -1623.7 3247.4 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5274 -0.6526 -0.2305 0.3031 7.7509
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 3.485e-14 1.867e-07
## Residual 1.438e-01 3.792e-01
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 9.5049 0.1778 53.46 <2e-16 ***
## TypeT 11.0292 0.2063 53.46 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(summary(m_em)$coeff[,1],4)## TypeP TypeT
## 9.505 11.030m_zm <- glmer(Zm ~ 0 + Type + (1|Mast), data=events[smp,], Gamma(link='identity'))
summary(m_zm)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: Zm ~ 0 + Type + (1 | Mast)
## Data: events[smp, ]
##
## AIC BIC logLik deviance df.resid
## 3635.8 3653.5 -1813.9 3627.8 620
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.0056 -0.7468 -0.1406 0.5875 3.9705
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 3.505e-14 1.872e-07
## Residual 4.277e-02 2.068e-01
## Number of obs: 624, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypeP 21.5480 0.2469 87.26 <2e-16 ***
## TypeT 22.7508 0.2607 87.26 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypeP
## TypeT 0.000signif(summary(m_zm)$coeff[,1],4)## TypeP TypeT
## 21.55 22.753.6 Un-filtered data
set.seed(12345)
smp <- names(which(table(as.character(datunf$Time))==4))
smp <- sample(smp, 250) # times 4 = 1000 records
smp <- which(datunf$Time %in% smp)
m_np <- glmer(NP_meas ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='log'))
summary(m_np)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: NP_meas ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 12922.2 12941.9 -6457.1 12914.2 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.9206 -0.5996 -0.2839 0.2625 10.4933
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 1.278e-16 1.130e-08
## Residual 1.055e+00 1.027e+00
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 5.25852 0.03613 145.5 <2e-16 ***
## TypeThi 5.74157 0.03613 158.9 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_np)$coeff[,1]),4)## TypePar TypeThi
## 192.2 311.6m_r <- glmer(R ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='log'))
summary(m_r)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: R ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 2471.5 2491.1 -1231.8 2463.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -0.5568 -0.4384 -0.2830 0.1104 18.5683
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.000 0.00
## Residual 2.756 1.66
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.16806 0.04353 3.861 0.000113 ***
## TypeThi 0.28237 0.04353 6.487 8.77e-11 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_r)$coeff[,1]),4)## TypePar TypeThi
## 1.183 1.326m_nd <- glmer(ND ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='log'))
summary(m_nd)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: ND ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 21721.8 21741.4 -10856.9 21713.8 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.5332 -0.6897 -0.1956 0.4718 13.7462
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 1.554e-14 1.246e-07
## Residual 3.884e-01 6.232e-01
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 9.784 0.026 376.3 <2e-16 ***
## TypeThi 10.416 0.026 400.6 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_nd)$coeff[,1]),4)## TypePar TypeThi
## 17750 33370m_z <- glmer(Z ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='log'))
summary(m_z)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: Z ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 6645.5 6665.2 -3318.8 6637.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.0737 -0.7527 -0.0638 0.6854 7.0290
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.08807 0.2968
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.11128 0.01326 234.7 <2e-16 ***
## TypeThi 3.17804 0.01326 239.8 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_z)$coeff[,1]),4)## TypePar TypeThi
## 22.45 24.00m_e <- glmer(E ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='log'))
summary(m_e)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( log )
## Formula: E ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 5531.3 5550.9 -2761.7 5523.3 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.4706 -0.7400 -0.2485 0.4481 6.2657
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.000 0.0000
## Residual 0.224 0.4733
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 2.19362 0.01886 116.3 <2e-16 ***
## TypeThi 2.33900 0.01886 124.0 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(exp(summary(m_e)$coeff[,1]),4)## TypePar TypeThi
## 8.968 10.370m_d10 <- glmer(D10 ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='identity'))
summary(m_d10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D10 ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## -2205.9 -2186.3 1107.0 -2213.9 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.8098 -0.6718 -0.1868 0.4154 4.5570
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 7.613e-07 0.0008726
## Residual 6.217e-02 0.2493323
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.500978 0.005314 94.28 <2e-16 ***
## TypeThi 0.240931 0.002656 90.72 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.047signif(summary(m_d10)$coeff[,1],4)## TypePar TypeThi
## 0.5010 0.2409m_d50 <- glmer(D50 ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='identity'))
summary(m_d50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D50 ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## -956.2 -936.6 482.1 -964.2 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.2929 -0.6064 -0.0976 0.4887 5.4000
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.000
## Residual 0.05856 0.242
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 0.803982 0.008372 96.04 <2e-16 ***
## TypeThi 0.530178 0.005521 96.04 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_d50)$coeff[,1],4)## TypePar TypeThi
## 0.8040 0.5302m_d90 <- glmer(D90 ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='identity'))
summary(m_d90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: D90 ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 104.5 124.1 -48.3 96.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.4958 -0.7027 -0.0964 0.4524 7.5111
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.000
## Residual 0.05381 0.232
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.25438 0.01217 103.1 <2e-16 ***
## TypeThi 1.12577 0.01092 103.1 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_d90)$coeff[,1],4)## TypePar TypeThi
## 1.254 1.126m_v10 <- glmer(V10 ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='identity'))
summary(m_v10)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V10 ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 796.9 816.6 -394.5 788.9 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.1700 -0.7007 -0.0996 0.5290 5.9505
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.0009373 0.03062
## Residual 0.0585776 0.24203
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 1.97175 0.05085 38.77 <2e-16 ***
## TypeThi 1.19854 0.04806 24.94 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.879signif(summary(m_v10)$coeff[,1],4)## TypePar TypeThi
## 1.972 1.199m_v50 <- glmer(V50 ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='identity'))
summary(m_v50)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V50 ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 1318.5 1338.1 -655.3 1310.5 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.3737 -0.5433 -0.0580 0.4113 4.9968
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.000
## Residual 0.03097 0.176
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 3.08547 0.02388 129.2 <2e-16 ***
## TypeThi 2.39234 0.01852 129.2 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_v50)$coeff[,1],4)## TypePar TypeThi
## 3.085 2.392m_v90 <- glmer(V90 ~ 0 + Type + (1|Mast), data=datunf[smp,], Gamma(link='identity'))
summary(m_v90)## Generalized linear mixed model fit by maximum likelihood (Laplace
## Approximation) [glmerMod]
## Family: Gamma ( identity )
## Formula: V90 ~ 0 + Type + (1 | Mast)
## Data: datunf[smp, ]
##
## AIC BIC logLik deviance df.resid
## 1966.3 1986.0 -979.2 1958.3 996
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -1.9609 -0.7136 -0.1612 0.5034 6.8924
##
## Random effects:
## Groups Name Variance Std.Dev.
## Mast (Intercept) 0.00000 0.0000
## Residual 0.02694 0.1641
## Number of obs: 1000, groups: Mast, 2
##
## Fixed effects:
## Estimate Std. Error t value Pr(>|z|)
## TypePar 4.20303 0.02897 145.1 <2e-16 ***
## TypeThi 4.21452 0.02905 145.1 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## TypePr
## TypeThi 0.000signif(summary(m_v90)$coeff[,1],4)## TypePar TypeThi
## 4.203 4.215We shall finally try fiting a heteroskedastic model, using glmmPQL from the MASS package. This function fits a Generalized Linear Mixed Effects Model with multivariate normal random effects, using Penalized Quasi-Likelihood.
# IT DOES NOT WORK
#library(nlme)
#library(MASS)
#glmmPQL(d50~0+Type, random=~1|Mast, family=Gamma(link='identity'),
# data=dat[smp,], weights=varIdent(form=~1|Type))